Advanced Lane Finding Project¶
The goals / steps of this project are the following:
- Compute the camera calibration matrix and distortion coefficients given a set of chessboard images.
- Apply a distortion correction to raw images.
- Use color transforms, gradients, etc., to create a thresholded binary image.
- Apply a perspective transform to rectify binary image ("birds-eye view").
- Detect lane pixels and fit to find the lane boundary.
- Determine the curvature of the lane and vehicle position with respect to center.
- Warp the detected lane boundaries back onto the original image.
- Output visual display of the lane boundaries and numerical estimation of lane curvature and vehicle position.
First, I'll compute the camera calibration using chessboard images¶
In [1]:
import numpy as np
import cv2
import glob
import matplotlib.pyplot as plt
import os
%matplotlib inline
In [2]:
def display_corners(fname, original_img, corners, ret):
corners_image = cv2.drawChessboardCorners(img, (9,6), corners, ret)
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(20,5))
ax1.imshow(original_img)
ax1.set_title('original_img:' + os.path.basename(fname))
ax2.imshow(corners_image)
ax2.set_title('corners_image:' + os.path.basename(fname))
# prepare object points, like (0,0,0), (1,0,0), (2,0,0) ....,(6,5,0)
objp = np.zeros((6*9,3), np.float32)
objp[:,:2] = np.mgrid[0:9,0:6].T.reshape(-1,2)
# Arrays to store object points and image points from all the images.
objpoints = [] # 3d points in real world space
imgpoints = [] # 2d points in image plane.
# Make a list of calibration images
cal_images = glob.glob('./camera_cal/calibration*.jpg')
# Step through the list and search for chessboard corners
for fname in cal_images:
img = cv2.imread(fname)
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
# Find the chessboard corners
ret, corners = cv2.findChessboardCorners(gray, (9,6),None)
# If found, add object points, image points
if ret == True:
objpoints.append(objp)
imgpoints.append(corners)
# Draw and display the corners
display_corners(fname, img, corners, ret)
Then undistort the image and visualize¶
In [3]:
cal_img_11 = cv2.imread(cal_images[11])
gray_cal_img_11 = cv2.cvtColor(cal_img_11, cv2.COLOR_BGR2GRAY)
ret, mtx, dist, rvecs, tvecs = cv2.calibrateCamera(objpoints, imgpoints,
gray_cal_img_11.shape[::-1], None, None)
def cal_undistort(img, objpoints = objpoints, imgpoints = imgpoints):
# Use cv2.calibrateCamera() and cv2.undistort()
undist = dst = cv2.undistort(img, mtx, dist, None, mtx)
return undist
def display_undistorted(image_path):
original_img = cv2.imread(image_path)
original_img = cv2.cvtColor(original_img, cv2.COLOR_BGR2RGB)
undistorted_img = cal_undistort(original_img)
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(10,5))
ax1.imshow(original_img)
ax1.set_title('original_img:' + os.path.basename(image_path))
ax2.imshow(undistorted_img)
ax2.set_title('undistorted_img:' + os.path.basename(image_path))
display_undistorted(cal_images[0])
display_undistorted(cal_images[11])
display_undistorted(cal_images[13])
Load the test images¶
In [4]:
test_images = glob.glob('test_images/*.jpg')
for img in test_images:
display_undistorted(img)
visualize an image for warping¶
In [5]:
straight_0 = cv2.imread(test_images[3])
straight_0 = cv2.cvtColor(straight_0, cv2.COLOR_BGR2RGB)
undistorted_straight_0 = cal_undistort(straight_0)
plt.imshow(undistorted_straight_0)
src = np.float32([[722, 470],[1120, 720],
[280,720],[570, 470]])
for point in src:
plt.plot(point[0], point[1],".")
Perspective Transform¶
Define the perspective transform function and visualize
In [6]:
def warpPerspective(img, to_birds_eye=True):
img_size = (img.shape[1], img.shape[0])
src = np.float32([[722, 470],[1120, 720],
[280,720],[570, 470]])
dst = np.float32([[920, 0], [920, 720],
[320,720],[320, 0]])
if to_birds_eye == False :
src, dst = dst, src
M = cv2.getPerspectiveTransform(src, dst)
return cv2.warpPerspective(img, M, img_size, flags=cv2.INTER_LINEAR)
def display_warped(image_path):
original_img = cv2.imread(image_path)
original_img = cv2.cvtColor(original_img, cv2.COLOR_BGR2RGB)
undistorted_img = cal_undistort(original_img)
birds_eye_img = warpPerspective(undistorted_img)
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(10,5))
ax1.imshow(original_img)
ax1.set_title('original_img:' + os.path.basename(image_path))
ax2.imshow(birds_eye_img)
ax2.set_title('birds_eye_img:' + os.path.basename(image_path))
for img in test_images:
display_warped(img)
Pipeline for images¶
Threshhold in color space for S and L and then the yellow and find the sobel binary in the x plane and angles close to center
In [7]:
def Get_S_Img(image):
hls = cv2.cvtColor(image, cv2.COLOR_RGB2HLS)
S = hls[:,:,2]
thresh = (100, 255)
binary = np.zeros_like(S)
binary[(S > thresh[0]) & (S <= thresh[1])] = 1
return binary
def Get_L_Img(image):
hls = cv2.cvtColor(image, cv2.COLOR_RGB2HLS)
L = hls[:,:,1]
thresh = 100
binary = np.zeros_like(L)
binary[(L > thresh)] = 1
return binary
def Get_Yellow(image):
threshold = 150
R = image[:,:,0]
G = image[:,:,1]
yellow = np.zeros_like(R)
yellow[(R > threshold) & (G > threshold)] = 1
return yellow
def abs_sobel_thresh(img, orient='x', thresh_min=0, thresh_max=255):
# Convert to grayscale
gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
# Apply x or y gradient with the OpenCV Sobel() function
# and take the absolute value
if orient == 'x':
abs_sobel = np.absolute(cv2.Sobel(gray, cv2.CV_64F, 1, 0))
if orient == 'y':
abs_sobel = np.absolute(cv2.Sobel(gray, cv2.CV_64F, 0, 1))
# Rescale back to 8 bit integer
scaled_sobel = np.uint8(255*abs_sobel/np.max(abs_sobel))
# Create a copy and apply the threshold
binary_output = np.zeros_like(scaled_sobel)
# Here I'm using inclusive (>=, <=) thresholds, but exclusive is ok too
binary_output[(scaled_sobel >= thresh_min) & (scaled_sobel <= thresh_max)] = 1
# Return the result
return binary_output
def dir_threshold(img, sobel_kernel=3, thresh=(0, np.pi/2)):
# Grayscale
gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
# Calculate the x and y gradients
sobelx = cv2.Sobel(gray, cv2.CV_64F, 1, 0, ksize=sobel_kernel)
sobely = cv2.Sobel(gray, cv2.CV_64F, 0, 1, ksize=sobel_kernel)
# Take the absolute value of the gradient direction,
# apply a threshold, and create a binary image result
absgraddir = np.arctan2(np.absolute(sobely), np.absolute(sobelx))
binary_output = np.zeros_like(absgraddir)
binary_output[(absgraddir >= thresh[0]) & (absgraddir <= thresh[1])] = 1
# Return the binary image
return binary_output
def Get_Sobel_IMG(img):
gradx = abs_sobel_thresh(img, 'x', 10, 200)
dir_binary = dir_threshold(img, thresh=(np.pi/6, np.pi/2))
sobel = np.zeros_like(gradx)
sobel[(gradx == 1) & (dir_binary == 1)] = 1
return sobel
def region_of_interest(img, vertices):
"""
Applies an image mask.
Only keeps the region of the image defined by the polygon
formed from `vertices`. The rest of the image is set to black.
"""
#defining a blank mask to start with
mask = np.zeros_like(img)
#defining a 3 channel or 1 channel color to fill the mask with depending on the input image
if len(img.shape) > 2:
channel_count = img.shape[2] # i.e. 3 or 4 depending on your image
ignore_mask_color = (255,) * channel_count
else:
ignore_mask_color = 255
#filling pixels inside the polygon defined by "vertices" with the fill color
cv2.fillPoly(mask, vertices, ignore_mask_color)
#returning the image only where mask pixels are nonzero
masked_image = cv2.bitwise_and(img, mask)
return masked_image
def process(image_path):
original_img = cv2.imread(image_path)
original_img = cv2.cvtColor(original_img, cv2.COLOR_BGR2RGB)
undistorted_img = cal_undistort(original_img)
birds_eye = warpPerspective(undistorted_img)
sobel = Get_Sobel_IMG(birds_eye)
s_img = Get_S_Img(birds_eye)
white = Get_L_Img(birds_eye)
yellow = Get_Yellow(birds_eye)
lanes = np.zeros_like(yellow)
lanes[((yellow == 1) | (white == 1)) & ((s_img == 1) | (sobel == 1))] = 1
#birds_eye_lanes = warpPerspective(lanes)
final_img = lanes
return [original_img, final_img]
def displayProcessed(image_path):
original_img, final_img = process(image_path)
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(10,5))
ax1.imshow(original_img)
ax1.set_title('original_img:' + os.path.basename(image_path))
ax2.imshow(final_img, cmap='gray')
ax2.set_title('pre-process:' + os.path.basename(image_path))
In [8]:
for img in test_images:
displayProcessed(img)
see histogram of procesed image¶
In [9]:
def displayHist(image_path):
original_img, processed_img = process(image_path)
histogram = np.sum(processed_img[processed_img.shape[0]//2:,:], axis=0)
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(10,5))
ax1.imshow(original_img)
ax1.set_title('original_img:' + os.path.basename(image_path))
ax2.plot(histogram)
ax2.set_title('histogram:' + os.path.basename(image_path))
for img in test_images:
displayHist(img)
Get left and right lanes¶
In [10]:
def GetLanesPoly(binary_warped):
# Assuming you have created a warped binary image called "binary_warped"
# Take a histogram of the bottom half of the image
histogram = np.sum(binary_warped[binary_warped.shape[0]//2:,:], axis=0)
# Create an output image to draw on and visualize the result
out_img = np.dstack((binary_warped, binary_warped, binary_warped))*255
# Find the peak of the left and right halves of the histogram
# These will be the starting point for the left and right lines
midpoint = np.int(histogram.shape[0]/2)
leftx_base = np.argmax(histogram[:midpoint])
rightx_base = np.argmax(histogram[midpoint:]) + midpoint
# Choose the number of sliding windows
nwindows = 9
# Set height of windows
window_height = np.int(binary_warped.shape[0]/nwindows)
# Identify the x and y positions of all nonzero pixels in the image
nonzero = binary_warped.nonzero()
nonzeroy = np.array(nonzero[0])
nonzerox = np.array(nonzero[1])
# Current positions to be updated for each window
leftx_current = leftx_base
rightx_current = rightx_base
# Set the width of the windows +/- margin
margin = 100
# Set minimum number of pixels found to recenter window
minpix = 50
# Create empty lists to receive left and right lane pixel indices
left_lane_inds = []
right_lane_inds = []
# Step through the windows one by one
for window in range(nwindows):
# Identify window boundaries in x and y (and right and left)
win_y_low = binary_warped.shape[0] - (window+1)*window_height
win_y_high = binary_warped.shape[0] - window*window_height
win_xleft_low = leftx_current - margin
win_xleft_high = leftx_current + margin
win_xright_low = rightx_current - margin
win_xright_high = rightx_current + margin
# Identify the nonzero pixels in x and y within the window
good_left_inds = ((nonzeroy >= win_y_low) & (nonzeroy < win_y_high) &
(nonzerox >= win_xleft_low) & (nonzerox < win_xleft_high)).nonzero()[0]
good_right_inds = ((nonzeroy >= win_y_low) & (nonzeroy < win_y_high) &
(nonzerox >= win_xright_low) & (nonzerox < win_xright_high)).nonzero()[0]
# Append these indices to the lists
left_lane_inds.append(good_left_inds)
right_lane_inds.append(good_right_inds)
# If you found > minpix pixels, recenter next window on their mean position
if len(good_left_inds) > minpix:
leftx_current = np.int(np.mean(nonzerox[good_left_inds]))
if len(good_right_inds) > minpix:
rightx_current = np.int(np.mean(nonzerox[good_right_inds]))
# Concatenate the arrays of indices
left_lane_inds = np.concatenate(left_lane_inds)
right_lane_inds = np.concatenate(right_lane_inds)
# Extract left and right line pixel positions
leftx = nonzerox[left_lane_inds]
lefty = nonzeroy[left_lane_inds]
rightx = nonzerox[right_lane_inds]
righty = nonzeroy[right_lane_inds]
# Fit a second order polynomial to each
left_fit = np.polyfit(lefty, leftx, 2)
right_fit = np.polyfit(righty, rightx, 2)
ploty = np.linspace(0, binary_warped.shape[0]-1, binary_warped.shape[0] )
left_fitx = left_fit[0]*ploty**2 + left_fit[1]*ploty + left_fit[2]
right_fitx = right_fit[0]*ploty**2 + right_fit[1]*ploty + right_fit[2]
return [left_fitx, right_fitx, ploty, left_fit, right_fit]
def getFilledImage(processed_img):
warp_zero = np.zeros_like(processed_img).astype(np.uint8)
color_warp = np.dstack((warp_zero, warp_zero, warp_zero))
left_fitx, right_fitx, ploty, left_fit, right_fit = GetLanesPoly(processed_img)
left_line = np.array(np.transpose(np.vstack([left_fitx, ploty])))
right_line = np.array(np.flipud(np.transpose(np.vstack([right_fitx, ploty]))))
poly_points = np.vstack([left_line, right_line])
cv2.fillPoly(color_warp, np.int_([poly_points]), [0,255, 0])
cv2.polylines(color_warp, np.int32([left_line]), isClosed=False, color=(255,0,0), thickness=15)
cv2.polylines(color_warp, np.int32([right_line]), isClosed=False, color=(255,0,0), thickness=15)
color_unwarp = warpPerspective(color_warp, to_birds_eye=False)
return [color_unwarp, left_fit, right_fit, left_fitx, right_fitx, ploty]
def weighted_img(img, initial_img, α=1., β=.4, λ=0.):
"""
`img` is the output of the hough_lines(), An image with lines drawn on it.
Should be a blank image (all black) with lines drawn on it.
`initial_img` should be the image before any processing.
The result image is computed as follows:
initial_img * α + img * β + λ
NOTE: initial_img and img must be the same shape!
"""
return cv2.addWeighted(initial_img, α, img, β, λ)
def radius(leftx, rightx, ploty):
# Define conversions in x and y from pixels space to meters
ym_per_pix = 30/720 # meters per pixel in y dimension
xm_per_pix = 3.7/700 # meters per pixel in x dimension
y_eval = np.max(ploty)
# Fit new polynomials to x,y in world space
left_fit_cr = np.polyfit(ploty*ym_per_pix, leftx*xm_per_pix, 2)
right_fit_cr = np.polyfit(ploty*ym_per_pix, rightx*xm_per_pix, 2)
# Calculate the new radii of curvature
left_curverad = ((1 + (2*left_fit_cr[0]*y_eval*ym_per_pix + left_fit_cr[1])**2)**1.5) / np.absolute(2*left_fit_cr[0])
right_curverad = ((1 + (2*right_fit_cr[0]*y_eval*ym_per_pix + right_fit_cr[1])**2)**1.5) / np.absolute(2*right_fit_cr[0])
return "Radius of curvature : %.2f m" % ((left_curverad+right_curverad)/2)
def center(leftx, rightx, image):
xm_per_pix = 3.7/700
midx = image.shape[1]/2
lanex = (leftx[-1] + rightx[-1]) /2
center_dist = (midx - lanex) * xm_per_pix
dir = "left"
if center_dist < 0 :
dir = "right"
center_dist = - center_dist
return "Car is " + dir + " of the center by: %.2f m" % (center_dist)
def displayLanes(image_path):
original_img, processed_img = process(image_path)
poly_img, left_fit, right_fit, left_fitx, right_fitx, ploty = getFilledImage(processed_img)
rad = radius(left_fitx, right_fitx, ploty)
cent = center(left_fitx, right_fitx, original_img)
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(14,10))
ax1.imshow(original_img)
ax1.set_title('original_img:' + os.path.basename(image_path))
final = weighted_img(poly_img, cal_undistort(original_img))
cv2.putText(final,rad , (40, 50), cv2.FONT_HERSHEY_SIMPLEX, 1.5, (255,255,255), 2, cv2.LINE_AA)
cv2.putText(final,cent , (40, 100), cv2.FONT_HERSHEY_SIMPLEX, 1.5, (255,255,255), 2, cv2.LINE_AA)
ax2.imshow(final)
ax2.set_title('lanes:' + os.path.basename(image_path))
for img in test_images:
displayLanes(img)
Process Video¶
In [11]:
def GetLanesFast(binary_warped, left_fit, right_fit):
nonzero = binary_warped.nonzero()
nonzeroy = np.array(nonzero[0])
nonzerox = np.array(nonzero[1])
margin = 100
left_lane_inds = ((nonzerox > (left_fit[0]*(nonzeroy**2) + left_fit[1]*nonzeroy +
left_fit[2] - margin)) & (nonzerox < (left_fit[0]*(nonzeroy**2) +
left_fit[1]*nonzeroy + left_fit[2] + margin)))
right_lane_inds = ((nonzerox > (right_fit[0]*(nonzeroy**2) + right_fit[1]*nonzeroy +
right_fit[2] - margin)) & (nonzerox < (right_fit[0]*(nonzeroy**2) +
right_fit[1]*nonzeroy + right_fit[2] + margin)))
# Again, extract left and right line pixel positions
leftx = nonzerox[left_lane_inds]
lefty = nonzeroy[left_lane_inds]
rightx = nonzerox[right_lane_inds]
righty = nonzeroy[right_lane_inds]
# Fit a second order polynomial to each
left_fit = np.polyfit(lefty, leftx, 2)
right_fit = np.polyfit(righty, rightx, 2)
# Generate x and y values for plotting
ploty = np.linspace(0, binary_warped.shape[0]-1, binary_warped.shape[0] )
left_fitx = left_fit[0]*ploty**2 + left_fit[1]*ploty + left_fit[2]
right_fitx = right_fit[0]*ploty**2 + right_fit[1]*ploty + right_fit[2]
return [left_fitx, right_fitx, ploty, left_fit, right_fit]
class Lanes():
def __init__(self):
self.fast = False
self.leftx = []
self.rightx = []
self.left_fit = []
self.right_fit= []
self.left_history = []
self.right_history = []
def getAverageLines(self, left_fitx, right_fitx):
frames = 20
#if left_fitx == None or left_fitx == [] :
# print("left is:", left_fitx)
#if right_fitx == None or right_fitx == [] :
# print("left is:", right_fitx)
if left_fitx != None and left_fitx != [] :
self.left_history += [left_fitx]
if right_fitx != None and right_fitx != [] :
self.right_history += [right_fitx]
if(len(self.left_history) > frames):
self.left_history.pop(0)
if(len(self.right_history) > frames):
self.right_history.pop(0)
new_left_fitx = np.zeros_like(left_fitx)
new_right_fitx = np.zeros_like(right_fitx)
for line in self.left_history:
new_left_fitx += line
new_left_fitx /= len(self.left_history)
for line in self.right_history:
new_right_fitx += line
new_right_fitx /= len(self.right_history)
return new_left_fitx, new_right_fitx
def getFilledImage(self, processed_img):
warp_zero = np.zeros_like(processed_img).astype(np.uint8)
color_warp = np.dstack((warp_zero, warp_zero, warp_zero))
if self.fast == True:
left_fitx, right_fitx, ploty, self.left_fit, self.right_fit = GetLanesFast(processed_img,
self.left_fit, self.right_fit)
else:
left_fitx, right_fitx, ploty, self.left_fit, self.right_fit = GetLanesPoly(processed_img)
self.fast = True
left_fitx, right_fitx = self.getAverageLines(left_fitx, right_fitx)
left_line = np.array(np.transpose(np.vstack([left_fitx, ploty])))
right_line = np.array(np.flipud(np.transpose(np.vstack([right_fitx, ploty]))))
poly_points = np.vstack([left_line, right_line])
cv2.fillPoly(color_warp, np.int_([poly_points]), [0,255, 0])
cv2.polylines(color_warp, np.int32([left_line]), isClosed=False, color=(255,0,0), thickness=15)
cv2.polylines(color_warp, np.int32([right_line]), isClosed=False, color=(255,0,0), thickness=15)
color_unwarp = warpPerspective(color_warp, to_birds_eye=False)
return [color_unwarp,left_fitx, right_fitx, ploty]
def video(self, original_img):
undistorted_img = cal_undistort(original_img)
birds_eye = warpPerspective(undistorted_img)
sobel = Get_Sobel_IMG(birds_eye)
s_img = Get_S_Img(birds_eye)
white = Get_L_Img(birds_eye)
yellow = Get_Yellow(birds_eye)
lanes = np.zeros_like(yellow)
lanes[((yellow == 1) | (white == 1)) & ((s_img == 1) | (sobel == 1))] = 1
final = lanes
poly_img,left_fitx, right_fitx, ploty = self.getFilledImage(final)
final2 = weighted_img(poly_img, cal_undistort(original_img))
rad = radius(left_fitx, right_fitx, ploty)
cent = center(left_fitx, right_fitx, original_img)
cv2.putText(final2,rad , (40, 50), cv2.FONT_HERSHEY_SIMPLEX, 1.5, (255,255,255), 2, cv2.LINE_AA)
cv2.putText(final2,cent , (40, 100), cv2.FONT_HERSHEY_SIMPLEX, 1.5, (255,255,255), 2, cv2.LINE_AA)
return final2
In [12]:
# Import everything needed to edit/save/watch video clips
from moviepy.editor import VideoFileClip
from IPython.display import HTML
lanes = Lanes()
project_output = 'proccessed_project_video.mp4'
clip1 = VideoFileClip("project_video.mp4")
project_clip = clip1.fl_image(lanes.video) #NOTE: this function expects color images!!
%time project_clip.write_videofile(project_output, audio=False)